DeleD Community Edition

[AWM Mars]

I would help with this project, not as a programmer, lol.. no.. but as a seasoned user, a immersive environments developer and an 'artist'.
_________________
Politeness is priceless when received, cost nothing to own or give, yet some cannot afford.

Checkout:
http://www.awm.mars.yourinside.com/
http://www.bccservices.co.uk
http://www.localtradecheck.co.uk

[chronozphere]

I would like to help out. I don't know C# but I do know a fair bit of JAVA so learning is not much of a problem. I'm especially interested in discussing and designing the foundations of DeleD.

I do have a limited amount of time for the project. I can contribute once in a while and keep the discussion going. We would probably need a few others before we can really start.

We should try to keep it as accessible as possible for developers who work in other languages. for example, it would be great if plugins could be written in other languages, or an integrated scripting language. We could also consider supporting other platforms through MONO.

[Jeroen]

I'm available for consultancy. As in, you can ask me questions on how things were built, and how I (now) think things should be build. As long as such questions don't take up too much of my precious little time, I'm very much ok with that.

Do note that the "final" product (if there is such a thing) should not be a rewritten DeleD, but something more than that. The aim should be to create a product which is easy to extend so that it is possible to evolve it into some sort of Scene Manager, which seems to be the next logical step as discussed previously in this here topic.

[Ewout]

[chronozphere]

[Starnick]

You guys should schedule a chat soon. I don't really know how much I could help out since I'm still busy with Tesla, but I wouldn't mind participating in design discussions.

Per what chronozphere said of accessibility for developers, I think that should be a low priority. However, switching to .NET (or Mono I suppose, I haven't used Mono before) you do have a lot of interop capabilities with P/Invoke, not to mention multiple languages that run on the CLR. With C# being so easy (my opinion, but it was for me) to get into, I don't think this would even be necessary. If someone wanted to have a plugin use C++ code that's already been written, they can just write a C# wrapper that uses P/Invoke.

Some big decisions would have to be made before moving forward too. Namely in regards to the renderer functionality.

If you want to continue with OGL, I would take a look at the OpenTK Framework:

http://www.opentk.com

There's another C# OGL library out there called Tao, but it seems OpenTK is the more active/complete/alive library out there. If I ever do write the OGL support for my engine that's the library I'm strongly considering to use.

If you want to change things up and use DirectX, then you obviously have XNA as a choice (just for rendering, content pipeline and game-related functionality can just be stripped out). Another choice would be SlimDX:

http://www.slimdx.org/

SlimDX (specifically their D3D10 API) is what I used for Tesla (for D3D9 I do plan on supporting XNA tho), and its really good - a lot better than XNA in my opinion if you just want DirectX functionality without all the sugar coating. But it's harder to use than XNA, however affords you more control especially if you don't want to just use D3D9.

Of course, you can always use both and have the user choose. That route probably wouldn't be recommended since it's a lot of work, because you'd probably spend more time writing an abstraction layer to run with either graphics API. I should know, because that's what I'm doing with Tesla. Of course you can always use existing technology out there that provides you that rendering support

One last thing I would think about is shader support. This is both a developer -and- a user consideration. If you forgo the fixed function pipeline (which you'd have to if you go with D3D10/11, XNA) you have to use shaders anyways. This comes at a cost though, especially if you support it for users since that makes things more complex.

It would also give deled users far more control over their models and scenes because it would give them the ability to create materials that can easily be used in their target application. Honestly I only ever used texture/color information from DeleD and just threw out the rest. But that comes at a cost because you'd have to be able to deal with custom shader uniforms and inputs that have to be modified via the GUI.

Of course you don't have to have user-support and just write your own code that the editor only ever uses. It would simplify things internally I think too.

If user support for shaders would be considered, you also have another consideration:

Should users be forced to use the renderer's shading langauge? (GLSL for OGL, HLSL (or using .FX files) for DX?). Or do you want to support Nvidia's Cg shading language since that can be supported by either?

[Ewout]

Ok I've started a very small design. I first checked out the functionality of DeleD. I made a small design how we could make tools more generic and start up a bit technically.



If I check some tools in Deled, I see that they could be seperated in groups (IPolygonTool etc). For example if you want to build in Deled functionality to move a vertex, you implement all the interfaces. If you want to make the Split function, you implement the IEdgeTool interface. It is possible to implement these interfaces. The core in the graphics take care that the correct vertex/polygon/edge etc is passed to the interface.

This is a first small start. Please give tips/comments/ideas. I haven't tought yet about things like how to make dynamic windows etc. What's the way you should do it Chronos?

[fik]

Good luck guys,although the direction this is going makes me sad.
I wish I had commital rights not just being able to submit patches, as would like to see Dele grow not a new project started.
I see some open source projects have seperate branches by diferent developers. Maybe this could be an option and bits liked could be put into official version. Still have lots of ideas, would like to add but must admit lack of action at sourceforge has rather disheartend my efforts for a while.

[Starnick]

@fik
This definitely would be a branched project and not be in the official release trunk unless if it actually gets to that point. Branching off is pretty common, even for big name software (Ogre3D has a fork called Axiom3D, which is C#).

@Ewout
The site's IRC channel would make a good place for a chat, Nocturn and I usually get lonely there .

Interfaces for tools is a good way to go. The Mesh/Meshparts is very XNA I see .

Dynamic windows...you mean like the viewports in the app, a la rendering to multiple windows? In DX you can use different swapchains/render targets for this, I'd imagine in OGL there's something similar.

Again I would say wanting to write an abstraction layer to support both OGL and DX is probably out of scope and would prevent you from working more on the application's core functionality. Its a lot easier said than done. But having a well designed rendering system would allow this to be incorporated in later possibly.

@Jeroen
How are vertices, polygons, and edges managed in DeleD currently? Just lists, or something more sophisticated?

[Jeroen]

A few remarks: